Application of the solute partitioning model to quantify Hofmeister ion effects on model processes and protein folding

Quantitative interpretation and prediction of Hofmeister salt effects on protein processes, including folding and crystallization, have been elusive goals of a century of research. We have developed a surface-bulk ion partitioning model, analogous to a model developed previously for the interpretation of solute effects on biopolymer processes, and applied it to analyze literature surface tension data for aqueous salt solutions. Single-ion partition coefficients are found to be independent of bulk salt concentration and additive for different salt ions. These partition coefficients provide a quantitative means of comparison with surface-sensitive spectroscopic measurements and with predictions of molecular dynamics simulations. In most cases, rank orders of partition coefficients for the air-water surface follow the conventional biopolymer Hofmeister series, but cations are shifted (toward exclusion) as compared with inferred values for interactions with protein surface.;The SPM is then applied to literature solubility/distribution data for hydrocarbons and model peptides. Both the hydration layer thickness and rank orders of anion and cation interactions with hydrocarbon surface are similar to those calculated for the air-water surface. Application of a course-grained surface area decomposition (e.g., nonpolar, polar amide) to the model peptide data allows determination of partition coefficients for amide surface. All salts investigated are found to accumulate at amide surface to a similar extent. Ion effects are independent and additive, allowing successful prediction of salt effects on model micelle formation and peptide solubility.;From the model compound analysis, we propose that the spectrum of Hofmeister salt effects observed for protein unfolding is due to the large amounts of nonpolar (hydrocarbon) surface exposed in this process. Salts that are only weakly excluded from nonpolar surface (e.g. GuHCl) are denturants because of their accumulation at amide surface. Since the vast majority of the surface exposed in unfolding is nonpolar or polar amide, we test the model compound predictions by applying the SPM to the thermodynamics of unfolding a small globular protein, monitored by circular dichroism spectroscopy. A preliminary separation of Coulombic and Hofmeister effects allows determination of Hofmeister interaction coefficients which are in good agreement with those predicted from the model compound analysis and coarse-grained DeltaASA decomposition.